ES2426321B1 - METHOD FOR IMPLEMENTING A MODE OF TRANSMISSION OF MULTIPLE INPUTS-MULTIPLE OUTPUTS - Google Patents

Abstract

Method for implementing a transmission mode of multiple inputs-multiple outputs. # In the method of the invention, said MIMO transmission mode comprises using a certain number of antennas in a user equipment and a given number of antennas in an access node. radio for communication between said user equipment and said radio access node, said communication is carried out by means of a communication channel, said communication channel represented by a channel matrix. # The method of the invention is characterized because it comprises de-mapping at least part of the coefficients of said channel matrix when said communication channel is degenerated using at least one additional antenna port on each of said radio access node and said user equipment in order to change the radiation pattern associated with transmission layers of said communication allowing the use of an enhanced MIMO transmission mode.

Description

Method to implement a transmission mode of multiple inputs-multiple outputs

Technical field

The present invention generally relates to a method for implementing a transmission mode of multiple inputs-multiple outputs, said MIMO transmission mode comprising using a certain number of antennas in a user equipment and a given number of antennas in a node of radio access for communication between said user equipment and said radio access node, said communication is carried out by means of a communication channel, said communication channel represented by a channel matrix and more particularly to a method that it comprises de-mapping at least part of the coefficients of said channel matrix when said communication channel is degenerated using at least one additional antenna port on each of said radio access node and said user equipment in order to change the pattern of radiation associated with transmission layers of said communication allowing the use of a powerful MIMO transmission mode born.

Prior art

Long-term evolution (LTE) is the next stage in cellular 3G systems, which basically represents an evolution of current mobile communications standards, such as UMTS and GSM [2]. It is a 3GPP standard that provides yields of up to 50 Mbps in uplink and up to 100 Mbps in downlink. It uses an adjustable bandwidth to scale from 1.25 to 20 MHz in order to adapt to the needs of network operators that have different bandwidth allocations. The LTE is also expected to improve spectral efficiency in networks, allowing carriers to provide more data and voice services over a given bandwidth.

Advanced LTE (LTE-A), an evolution of LTE, is normalizing in release 10 of LTE and later. It is aimed at meeting advanced IMT requirements, whose capabilities go beyond those of IMT-2000 and include enhanced peak data transmission rates to support advanced applications and services (100 Mbps for high mobility, and 1 Gbps for low mobility ) [3].

The use of multiple antenna technology allows us to take advantage of the space field as another new dimension. This becomes essential in the search for superior spectral efficiencies. Multiple antennas can be used in a variety of ways, based primarily on three fundamental principles [4]:

-

Diversity gain Use of the spatial diversity provided by the multiple antennas to improve the robustness of the transmission against multipath fading.

-

Network Gain Energy concentration in one or more directions given by precoding or beam modeling.

-

Gain of spatial multiplexing. Transmission of multiple signal flows to a single user over multiple spatial layers created by combinations of the available antennas.

Spatial multiplexing is the only scheme that achieves an increase in the capacity of available resources (time, frequency), multiplying the performance by a number that, under ideal conditions, is equal to the minimum number of transmitting and receiving antennas. This number is called the channel range, and depends strongly on the radio conditions and also on the geometric arrangement of both TX and RX antennas.

The spatial multiplexing scheme is the only one useful when the channel is not degenerated, that is, when the channel range is a maximum. This happens when all the components of the channel matrix are uncorrelated. Assuming that you have a SU-MIMO system with N transmit antennas and M receive antennas, the channel matrix is as follows:

 H H

00 0, N -1

H = [[:: JJ [J

H H

 M -1.0 M -1, N -1

,

where hij represents the complex channel coefficient between the transmitting antenna j and the receiving antenna i, as shown in Figure 1. If all channel components are de-correlated, the range of this matrix will be equal to the minimum of (N , M). However, if the propagation environment "observed" by the radiated beams is very similar, the channel becomes degenerated with a range that is less than min (N, M). If the range becomes one, spatial multiplexing becomes impractical and the system must resort to a scheme of

ES 2 426 321 A2

diversity.

It is important to note that channel degeneration cannot be compensated by precoding, regardless of the actual scheme used as defined for LTE and LTE-A: precoding schemes based on codebook for closed loop MIMO (mode 4 of LTE transmission) , or CDD (cyclic delay diversity) for open-loop MIMO (mode 3 of LTE transmission). This is due to the fact that both precoding and CDD can be considered an additional matrix multiplication operation in the H-channel matrix: if X is the vector containing the complex modulated symbols of each layer, and Y the complex vector after the precoding, the operation can be expressed [5]:

-

For TM 3 (OL-MIMO): Y = W (i) · D (i) · U · X, where W (i) is a precoding matrix, D (i) a diagonal matrix containing the applied cyclic delay, and U a square matrix dependent on the number of layers.

For TM 4 (CL-MIMO): Y = W (i) · X, where W (i) is a precoding matrix.

The precoded vector passes through the H channel, and the reception vector can be expressed as H · Y. Therefore, taking into account that for two matrices A and B given any one the following property [6] applies:

range (AB) � min (range (A), range (B)),

it is clear that if the matrix H is degenerated the precoding operation will not improve the demodulation process.

Due to this, the normal operating mode referring to SU-MIMO in LTE-A will be shown in the figure

2: eNode B receives periodic or non-periodic reports from the UEs through the PUCCH or PUSCH uplink channels (2a), including the range indicator (RI) that reports the downlink channel range as observed by the EU. Additionally, eNode B estimates the uplink channel range through, for example, uplink SRS transmissions. Based on these channel ranges, and on the channel quality indicators (CQI), eNode B decides whether or not to use a spatial multiplexing scheme in both uplink and downlink directions:

-

If the CQI is lower than some predefined thresholds (2b), the system uses a one-layer transmission scheme (2c).

-

If the CQI is higher than the threshold mentioned above, eNode B selects several layers for the transmission that is always less than or equal to the corresponding range. If the range is one (2d), the system again uses a one-layer transmission scheme to avoid degeneration of the channel matrix. Otherwise, the system takes full advantage of spatial multiplexing (2e).

Channel matrix degeneration is a situation that must be avoided in order to achieve the expected yields of LTE and LTE-A, because diversity schemes cannot take advantage of the spatial multiplexing of having multiple antennas.

There are also solutions to overcome this problem. For example, [8] describes a system for use in MIMO vehicle systems, which has several directional sub-networks mounted on different faces of the vehicle. As each sub-network will experience different channel conditions, based on the channel range or other appropriate measure the system can select the sub-network with the best expected performance. The drawback of this solution is that it is only suitable for vehicular conditions, because a large amount of space is needed to locate the different sub-networks, making the system more complex and expensive.

Another solution is proposed in [9], in which the configuration and radiation / polarization properties of the network are adaptively modified by altering the physical structure of the antenna components. This solution involves complicated reconfigurable antenna elements (with micro-electromechanical actuators), and a processing unit to find the optimal arrangement. Therefore, it may not be suitable for low cost user equipment.

[10] another invention is described, in which a multi-beam antenna employs two different network antennas. A maximum radiation direction of a combined beam of the second network antenna is oriented to a direction corresponding to a null point of the combined beam of the first network antenna. By doing this, the two antenna beams have a fairly low correlation and can be arranged in nearby locations. The drawback of this invention is that the beams are directed towards fixed directions, so that there is no possibility of adaptively following the channel characteristics. In addition, the two beams are achieved using a total of at least four antennas, which can increase the overall cost and complexity.

Finally, in [11] the approach is to configure a network of antennas by dividing it into sub-networks, based on the antenna correlation matrix, so that the correlations between antennas within each sub-network are superior to the correlations between antennas They belong to different sub-networks. Weights of

ES 2 426 321 A2

beam modeling for each antenna within the sub-networks, and a MIMO mechanism is also applied between the sub-networks. This scheme has the limitation of requiring a high number of antennas to form the sub-networks (at least two for each transmission layer), as well as the calculation of the correlation matrix by, for example, a calibration procedure, or by Feedback signals from mobile terminals.

Description of the invention

It is necessary to offer an alternative to the state of the art that covers the gaps found therein, particularly in relation to the lack of proposals that allow the use of spatial multiplexing techniques when the communication channel is degenerated.

For this, the present invention provides a method for implementing a transmission mode of multiple inputs-multiple outputs, said MIMO transmission mode comprising using a certain number of antennas in a user equipment and a given number of antennas in an access node radio for communication between said user equipment and said radio access node, said communication is carried out by means of a communication channel, said communication channel represented by a channel matrix.

Contrary to the known proposals, the method of the invention, in a characteristic way comprises de-correlating at least part of the coefficients of said channel matrix when said communication channel is degenerated using at least one additional antenna port in each of said radio access node and said user equipment in order to change the radiation pattern associated with transmission layers of said communication allowing the use of an enhanced MIMO transmission mode.

Other embodiments of the method of the first aspect of the invention are described according to the appended claims 2 to 15, and in a subsequent section related to the detailed description of various embodiments.

Brief description of the drawings

The advantages and features above and others will be more fully understood from the following detailed description of embodiments, with reference to the attached drawings (some of which have already been described in the prior art section), which should be considered in an illustrative and non-limiting manner, in which:

Figure 1 shows a current SU-MIMO system, with N transmission antennas and M reception antennas.

Figure 2 shows a current mode of operation of a SU-MIMO system in an LTE-A deployment.

Figure 3 shows the proposed mode of operation for a new MIMO transmission node with enhanced spatial multiplexing, according to an embodiment of the present invention.

Figure 4 shows a possible configuration for transmission considering an example of two monopole antennas, according to an embodiment of the present invention.

Figure 5 shows a schematic arrangement of two monopole antennas together with an additional central antenna seen from a perspective from above, according to an embodiment of the present invention.

Figures 6 and 7 show the radiation pattern resulting from each transmission layer for the case of d = λ / 2 when the appropriate values are chosen for the relative phases of two monopole antennas together with an additional central antenna in order to have nulls along the directions Φ = ± π / 4, ± 3 π / 4 in the azimuth plane, according to an embodiment of the present invention.

Figures 8 and 9 show the radiation pattern resulting from each transmission layer in the case of d = λ / 2 when the appropriate values are chosen for the relative phases of two monopole antennas together with an additional central antenna in order to have nulls along the directions Φ = 0, ± π / 2 in the azimuth plane, according to an embodiment of the present invention.

Figure 10 shows a schematic arrangement of pairs of linear cross polarization antenna elements together with a panel antenna with its pairs of cross polarization elements, according to an embodiment of the present invention.

Figure 11 shows a typical sector antenna radiation pattern.

Figures 12 and 13 show the radiation pattern resulting from each transmission layer for the case of d = λ / 2 when the appropriate values are chosen for the relative phases of pairs of linear cross polarization antenna elements together with an antenna of panel with its pairs of cross polarization elements in order to have nulls along the directions Φ = ± π / 4, ± 3 π / 4 in the azimuth plane, according to an embodiment of the present invention.

ES 2 426 321 A2

Figures 14 and 15 show the radiation pattern resulting from each transmission layer for the case of d = λ / 2 when the appropriate values are chosen for the relative phases of pairs of linear cross-polarized antenna elements together with an antenna of panel with its pairs of cross polarization elements in order to have nulls along the directions Φ = 0, ± π / 2 in the azimuth plane, according to an embodiment of the present invention.

Figures 16 and 17 show the radiation pattern resulting from each transmission layer in the case of d <λ / 2 when the appropriate values are chosen for the relative phases of two monopole antennas together with an additional central antenna in order to have nulls along the directions Φ = ± π / 4, ± 3 π / 4 in the azimuth plane, according to an embodiment of the present invention.

Figures 18, 19 and 20 show the radiation pattern resulting from each transmission layer in the case of d <λ / 2 when the appropriate values are chosen for the relative phases of two monopole antennas together with an additional central antenna in order of having nulls along the directions Φ = 0, ± π / 2 in the azimuth plane, according to an embodiment of the present invention.

Figure 21 shows a schematic arrangement of two monopole antennas together with an additional antenna forming an equilateral triangle seen from a perspective from above, according to an embodiment of the present invention.

Figures 22 and 23 show the radiation pattern resulting from each transmission layer for the values φ 1 = φ 2 = 0 in the case of a triangular network, according to an embodiment of the present invention.

Figures 24 and 25 show the radiation pattern resulting from each transmission layer for the values φ 1 = φ 2 = π / 2 in the case of a triangular network, according to an embodiment of the present invention.

Figure 26 shows a schematic arrangement of a transmitter with N monopole antennas and N / 2 additional interlaced antenna ports, according to an embodiment of the present invention.

Figure 27 shows the generic case for cross polarization antennas, according to an embodiment of the present invention.

Figure 28 shows a possible block diagram of the enhanced MIMO transmission mode proposed in the present invention.

Detailed description of various embodiments

The present invention proposes to introduce a new MIMO mode of enhanced spatial multiplexing (SM) transmission, which will make sense when the channel range is less than the minimum number of transmit and receive antennas (ie there is no full range transmission) , as shown in figure 3.

Upon receipt of the CQI and RI of the UEs (3a), the system evaluates the channel range (3b). If the transmission is full-range, it works in one of the usual spatial multiplexing (3d) modes, either open loop or closed loop. Otherwise, the system operates in an "enhanced spatial multiplexing transmission mode" which is the object of the present invention (3c).

In order to clarify the proposed enhanced SM transmission mode, an SU-MIMO 2x2 system from LTE-A will be assumed. The basic idea proposed in this invention is to use an additional antenna port, or several additional ones, for both transmission and reception in order to differentiate the radiation patterns of the different transmission layers. This is expected to enhance the channel characteristics so that the hij array elements are de-correlated.

It is not necessary that the new antenna ports be connected to any additional power amplifier (in transmission) or LNA (in reception), apart from those present in a typical MIMO system. Instead, they are intended to form networks of "virtual" antennas along with the other antenna elements. In transmission, each new antenna element will be fed with a linear combination of the signals that constitute the transmission layers. On reception, the RF signals coming from each antenna are combined again linearly before entering the LNA.

Linear combinations can be obtained by typical passive elements such as for example dimmers, dividers, circulators, etc. Therefore, the total transmission (and reception) power will be the same as with a normal MIMO system.

Figure 4 represents a possible configuration for the transmission considering an example of two monopole antennas (however, the present invention does not exclude any type of antenna). In addition to antennas TX1 and TX2 (each corresponding to a transmission layer), fed with complex signals T1 and T2

ES 2 426 321 A2

respectively, a third antenna TX3 is located between the other two, and is fed with a linear combination of signals T1 and T2.

The third antenna serves the purpose of creating two networks of “virtual” antennas, one for each transmission layer. The first network is composed of the antenna TX1 and part of the new antenna (central), the part corresponding to the signal a · T1, being for reasons of simplicity a complex unit factor in the form exp (iφ1). The second network is composed of the TX2 antenna and the part of the central antenna corresponding to the b · T2 signal, with b being in the exp (iφ2) form.

The purpose of these networks is to change the radiation pattern associated with the two transmission layers. It is assumed that, in the absence of the third central antenna, the H-channel matrix is unique, so that its range is equal to one. Considering first the part of the channel that comprises the transmitting antennas and their surroundings, the degeneration of the H-channel matrix means that the channels "seen" by the radio signals coming from the two antennas are highly correlated. This situation appears, for example, when a significant LOS (line of sight) component is present between the transmitter and the receiver, and is particularly harmful. One means to avoid this situation is to change the radiation pattern in such a way that the channel coefficients are different for both transmission layers.

It is possible to choose coefficients a, b in order to achieve nulls or peaks along different directions of space, thus eliminating the degeneration of the channel matrix. The receiver does not need to know the actual coefficients; The transmitter can also vary its values dynamically in order to find the most suitable combination (for example following precalculated patterns that enforce nulls or peaks along predefined directions of space).

On the reception side there is an analogous situation. By adjusting different phase factors in the signals received by each antenna, the reception radiation pattern will change creating new nulls and peaks that, together with those created on the transmission side, are expected to enhance the channel characteristics.

In order to apply the same mechanism in both TX and RX, eNode B will inform the UE of the proposed enhanced spatial multiplexing transmission mode, by means of an appropriate control message that will not be discussed here.

The proposed invention is equally valid for spatial multiplexing modes both open loop and closed loop. Next, two new transmission modes can be considered: "enhanced open loop spatial multiplexing" and "enhanced closed loop spatial multiplexing", which will be the "enhanced" analogs to modes 3 and 4 of LTE-A transmission, respectively.

It is important to emphasize that no complementary power amplifier is required for the TX3 central antenna. Instead, only a passive combination network is needed, because the total transmission power must be the same as without the complementary antenna. However, a proprietary solution may employ an additional power amplifier for this antenna in order to improve the gain, the noise figure, etc., but this is not required for the present invention.

The concept described in the present invention can be generalized in the case of having an arbitrary number of transmitting and receiving antennas. However, with respect to an LTE-A system, a 2x2 SU-MIMO configuration will be considered (later it will be generalized to NxM antennas).

• Example with linearly copolarized antennas

Figure 6 shows a schematic arrangement of two monopole TX1 and TX2 antennas together with an additional central antenna TX3, as seen from a perspective from above. The antenna elements TX1 and TX2 form the usual MIMO 2x2 combination (in this case being linearly copolarized), and an additional TX3 port has been added in the middle thereof.

The complex RF signals T1 and T2 feed the corresponding antenna ports TX1 and TX2, and the TX3 port is fed with a linear combination:

i �T

T = exp () ��T �exp () i�

3 11 22

The separation between the antenna elements is d and is assumed to be greater than or equal to π / 2. This condition is usually met in eNode B in order to keep the antenna correlation to a minimum. Having a high separation is beneficial for the proposed invention.

The resulting combination results in two networks of “virtual” antennas, one for each transmitted layer:

-

 Network formed by the radiated signals T1 + exp (iφ1) T1, coming from the antenna ports TX1 and TX3.

ES 2 426 321 A2

-

 Network formed by the radiated signals T2 + exp (iφ2) T2, coming from the antenna ports TX2 and TX3.

These two networks can be analyzed with the help of the so-called network factor of the structure, which for a linear network of M elements is defined as [7]:

M -1

AF (c) = II exp i (-kr)

m mm

m = 0.

where:

Im are the streams of the network elements;

2�

k = u

 being a unit vector in the direction of propagation;

rm are the position vectors of the network elements;

φm are the relative phases of the network elements.

Ф is the angle measured in the azimuth plane.

The overall radiation pattern is the product of the network factor and the radiation pattern of the individual antenna elements (assuming they are all identical). In the case of monopoles the radiation is omnidirectional in the azimuth (plane Ф), so that the resulting diagram is identical to the network factor, but with other antennas this may not be the case.

In the proposed scenario, taking the origin as the center of each of the networks considered, the two normalized network factors become:

kd cos c-1

AF () c = cos [J

TX1-TX 3

For TX1 + TX3:

kd cos c 2

AF () c = cos [J

TX 2-TX 3

For TX2 + TX3:

Multiple possibilities appear for the choice of phases φ 1 and φ 2. The present invention is equally valid regardless of the actual phases, which will be specific to the implementation.

-

 Nulls along the directions

Ф = ± π / 4, ± 3π / 4

Assuming that it is desired to have different radiation nulls for the T1 and T2 signals, specifically in the π / 4 and 3 π / 4 directions, in order to favor the radio paths that are to be de-correlated. Forcing, for example, that the TX1-TX3 network be null in the 3 π / 4 direction, and that the TX2-TX3 network be null in the π / 4, find the values for the phases:

d

1 = � [-1- 2 J

,

d

2 = � [1- 2 J

.

Figures 6 and 7 show the resulting radiation pattern of each transmission layer for the case d = λ / 2. It can be seen that the radiation patterns are somewhat deviated.

It is important to note that, when the antennas are separated a distance equal to or greater than half a wavelength, multiple lobes appear. This situation should be avoided in typical networks, in which it is important to obtain few secondary lobes, but for the present invention it is beneficial: the objective is not to concentrate the energy along a certain direction, but to change the radiation pattern of each layer of transmission and

ES 2 426 321 A2

Disperse energy in a different way for both layers. The appearance of high secondary lobes enhances the dispersion of energy and therefore the correlation properties of the channel matrix.

-

 Nulls along the directions Ф = 0, ± π / 2

This solution is obtained by forcing the TX1-TX3 network to be zero along 0º, and TX2-TX3 along ± π / 2. The relative phases are:

2d

1 = [� -1J

=

The resulting radiation patterns are shown in Figures 8 and 9 for d = λ / 2.

This example allows a clear separation of the maximum gain directions.

Other multiple configurations are possible by selecting different values of the relative phases, and this invention does not prefer the use of any value with respect to another. The eNode B may change and test various values of the relative phases without informing the user terminals, thus allowing an adaptive search of the optimal configuration for each case.

• Example with linear cross polarization antennas

This is the case most commonly found in cellular systems, because it achieves a low correlation between antenna elements. The analysis is identical, except that in this case the antenna radiation pattern has the typical sectorial shape, which influences the resulting global radiation pattern.

The antenna configuration is shown in Figure 10. Pairs of cross-polarized antenna elements are arranged forming a panel antenna, called "TX1 + TX2" to maintain the same terminology as in the linear case. The additional antenna element, TX3, is really another panel antenna with its pairs of cross polarization elements. Apart from the different geometry, the concept is analogous to the linear case; however, the difference is that the antennas TX1 and TX2 have orthogonal polarizations (which improves the correlation properties), and therefore the elements in TX3 are also orthogonal.

Different radiation patterns can now be obtained without using any power splitter, because the antenna elements are physically separated. Now virtual networks “TX1-TX3” and “TX2-TX3” are formed by the elements in both antennas that share the same polarization, as shown in Figure 10.

Figure 11 shows a typical 120 ° sector antenna radiation pattern in the azimuth plane (the one of interest for the present invention).

Due to the geometric arrangement, the two virtual networks are identical in this case and the corresponding network factors have the form:

kd cos c-1

AF () c = cos [J

TX1-TX 3

For TX1 + TX3:

kd cos c-2

AFTX 2-TX 3 () c = cos [J

For TX2 + TX3:

The analysis is therefore analogous to the case of monopole antenna, simply changing the sign of φ 2.

The overall radiation pattern, taking into account the antenna gain, is shown in Figures 12 and 13 for the case of nulls along the directions Ф = ± π / 4, ± 3 π / 4 and in Figures 14 and 15 for the case of nulls along the directions Ф = 0, ± π / 2.

It is clear that multiple combinations are possible in order to force different propagation directions for the two transmission layers. Again, any value can be used depending on the actual implementation, hardware limitations, granularity of the applicable phases, and so on.

• Enhanced SM 2x2 transmission mode, UE side

ES 2 426 321 A2

In the user terminal it is more difficult to have half wavelength antenna separations; therefore it is worth examining the case of a lower antenna separation. The radiation pattern of each antenna is usually omnidirectional in the azimuth, therefore the network factor coincides with the global radiation pattern in this plane.

Figures 16 and 17 show the corresponding diagrams obtained with a separation of 0.2 λ. It is clear that no secondary maxima appear in this case, only small lobes that point in the opposite direction of the main lobe.

In the case of nulls along the addresses Ф = 0, ± π / 2 there are two possible solutions for the TX1-TX3 network, each of them clearly pointing in opposite directions. The TX2-TX3 network has two identical lobes pointing to 0 and π, as shown in Figures 18, 19 and 20.

It is clear that these results are not as attractive as those in the case of the transmitter, due to the low antenna separation. To overcome this, another possible geometric arrangement for the receiving antennas is explained below.

-

 Triangular net

This configuration is shown in Figure 21, in which the antennas form an equilateral triangle.

The third antenna TX3 is now located at the apex of a triangle, the other two antennas being located in the other two vertices. This solution is more compact and naturally gives rise to different propagation directions for the two transmission layers, despite occupying a volume somewhat higher than the linear case.

The network factors for this configuration, taking into account that Φ ’= Φ– π / 3, are:

kd cos (c- / 3) -1

AFTX1-TX 3 () c = cos [J

For TX1 + TX3:

kd cos (c / 3) 2

AFTX 2-TX 3 () c = cos [J

For TX2 + TX3:

Due to the geometric arrangement the two networks are directed in different directions even with a zero phase, that is φ 1 = φ 2 = 0. Figures 22 and 23 show the radiation patterns in this case.

Setting φ 1 = φ 2 = π / 2 gives the patterns shown in figures 24 and 25.

Again there are multiple possibilities depending on the relative phases. The possibilities of obtaining hi-de-linked channel elements and, therefore, full-range reception are high in this case, due to the geometric arrangement of the antennas.

• Enhanced SM NxM transmission mode

It is easy to generalize the concept described in the present invention to a MIMO system with N transmission antennas and M reception antennas. The following figure illustrates the case of transmitter with N monopole antennas.

Together with the usual N transmission antennas TX1, TX2, ..., TXN, there are N / 2 additional interlaced antenna ports, called TX12, TX34, ..., TXN-1, N, so that N / 2 virtual networks are formed with each pair i, i + 1 (if N is odd the last antenna will remain unpaired). To simplify the analysis, all the antennas in the figure are equidistant apart a distance d, but in practice it may not be the case, depending on the particular geometry considered. The antennas TX1, TX2, ..., TXN are fed with signals T1, T2, ..., TN, and the new antenna elements are fed with linear combinations thereof:

exp () iT exp () iT

11 22

exp () iT exp () iT

33 44

() T exp iT

exp i ()

N-1 N-1 NN

ES 2 426 321 A2

Each pair of antennas gives rise to two virtual networks, one for each transmission layer, in a way that is similar to 2x2. Each transmission layer will therefore have a different radiation pattern, so that the correlation properties of the H-channel matrix are enhanced.

At reception the scheme is analogous, and the signals received from the antennas must be combined before entering the normal demodulation process, so that the desired beam modeling is applied.

In case of considering sector antennas, the diagram is the one represented in Figure 27

With this arrangement the networks are fed separately for each component, for example exp (iφ1) · T1 and exp (iφ2) · T2. However, the analysis is identical in this case.

Figure 28 shows a preferred embodiment of the proposed invention:

The system assesses the need to switch to the proposed enhanced SM transmission mode (block 301), based on EU reports (CQI, PMI and RI) and uplink probing (SRS) signals. If the conditions are not suitable for the proposed enhanced transmission mode (for example if the channel SNR is not high enough, or the channel range is a maximum) the system operates normally in any of the open loop transmission modes or usual closed loop (block 302). On the contrary, if the channel SNR is greater than a predefined threshold and the channel is not operating at full range, the system activates the enhanced SM transmission mode by means of an appropriate control message (block 303). This in turn activates additional antenna ports as explained above.

The selection of the relative phases to feed the additional antennas is done in block 304. The actual chosen values are completely dependent on the implementation and can be based on a set of predefined patterns (for example, forcing nulls or peaks along certain addresses in space), as discussed above.

The choice of relative phases follows an iterative procedure in which the system applies the different phases (block 305) and evaluates whether the channel is full range (block 306). If the channel is still degenerated, the system can change the phases in order to find the optimal values. Eventually, if after several iterations the channel characteristics do not improve, the system may resort to a normal SM mode as in block 302.

It is important to emphasize that the new enhanced mode does not interfere with legacy UEs, that is, the UEs that do not implement the proposed invention. If a legacy UE receives a control message indicating the use of an enhanced transmission mode, it will discard it and consider it identical to the corresponding normal transmission mode; however, eNode B can still apply the proposed invention in order to improve the channel properties on the TX side. Therefore the control message will not interfere with legacy UEs, and will be discarded by them thus operating in a normal SM mode. The control message is not described in the present invention.

-

 Advantages of the invention

The proposed invention addresses the problem of channel matrix degeneration in MIMO systems, which imposes a serious limitation on the use of multi-antenna techniques as described for advanced LTE. Other common solutions, such as adaptive networks, are expensive and difficult to implement. The techniques described by 3GPP to enhance the channel correlation characteristics (such as precoding or CDD) are not effective when the transmission is not full-range.

The advantage over other solutions is that the proposed LTE-A spatial multiplexing scheme can enhance the channel characteristics in both eNode B and in the UE, thus addressing the problem of end-to-end correlation. This MIMO mode of enhanced SM will be signaled from eNode B, and UEs with enhanced SM capability may apply the proposed scheme, while legacy UEs will not be affected by it.

Other solutions are more complicated and expensive, and yet do not deal with the correlation problem on both sides (TX and RX). In [8] the described invention is only valid for vehicular conditions, while the proposed invention is equally suitable for any environment. The solution in [9] is much more complicated to manage and build, making it not suitable for low cost UEs, while the present invention is particularly simple and economical. In [10] the described multi-beam antenna has the disadvantage of pointing towards fixed directions of space, thus making it non-adaptive, while the proposed invention allows an adaptive and proprietary search of the optimal configuration. In [11] a correlation matrix must be obtained, thereby complicating the operation of the system, and the number of antennas required is also greater than that of the present invention.

The proposed enhanced transmission mode is a simple and economical way to improve the channel properties, implementing different channel coefficients for the transmission layers in order to

ES 2 426 321 A2

enhance correlation characteristics. Since no additional power amplifiers or LNAs are needed for the complementary antenna elements, there is a minimal increase in complexity in eNodes B and the UEs.

The operation of the enhanced transmission mode under channel conditions that would otherwise not be suitable for spatial multiplexing (such as LOS) results in an increase in performance for the user, leading to superior cellular performance and an increase in revenue . In addition, the quality of experience is also enhanced, as users can receive multiple spatial streams for the same radio conditions.

A person skilled in the art may introduce changes and modifications to the described embodiments without departing from the scope of the invention as defined in the appended claims.

ACRONYM

3GPP

Third Generation Partnership Project, third generation partnership project

AF

Array Factor, network factor

CDD

Cyclic Delay Diversity, cyclic delay diversity

CQI

Channel Quality Indicator, channel quality indicator

GSM

Global System for Mobile Communications, global system for mobile communications

IM T

International Mobile Telecommunications, international mobile telecommunications

LNA

Low Noise Amplifier, low noise amplifier

THE

Line of Sight, line of sight

LTE

Long-Term Evolution, long-term evolution

LTE-A

Long-Term Evolution Advanced, advanced long-term evolution

MIME

Multiple Input - Multiple Output, multiple inputs-multiple outputs

OL-MIMO

Open-Loop MIMO, open-loop MIMO

CL-MIMO

Closed-Loop MIMO, closed loop MIMO

PMI

Precoding Matrix Indicator, precoding matrix indicator

PUCCH

Physical Uplink Control Channel, physical uplink control channel

PUSCH

Physical Uplink Shared Channel, physical uplink shared channel

RI

Rank Indicator, range indicator

RX

Receiver, receiver

YE

Spatial Multiplexing, spatial multiplexing

SNR

Signal to Noise Ratio, signal to noise ratio

MR

Sounding Reference Signal, polling reference signal

SU-MIMO

Single-User Multiple Input - Multiple Output, multiple inputs-multiple outputs of one

single user

TM

Transmission Mode, transmission mode

TX

Transmitter, transmitter

EU

User Equipment, user equipment

UMTS

Universal  Mobile Telecommunication  System, system  telecommunication  mobile

universal

ES 2 426 321 A2

BIBLIOGRAPHY

[1] Patent definition. http://en.wikipedia.org/wiki/Patent.

[2] 3GPP TS 36.300, Evolved Universal Terrestrial Radio Access (E-UTRA) and Evolved Universal Terrestrial Radio Access Network (E-UTRAN); Overall Description, Stage 2 (Release 8)

5 [3] http://www.3gpp.org/LTE-Advanced

[4] S. Sesia, I. Toufik, M. Baker (editors), "LTE, the UMTS Long Term Evolution: From Theory to Practice", John Wiley & Sons, 2009

[5] 3GPP TS 36.211, Evolved Universal Terrestrial Radio Access (E-UTRA); Physical channels and modulation (Release 10)

10 [6] http://en.wikipedia.org/wiki/Rank_(linear_algebra)

[7] C.A. Balanis, “Antenna Theory: Analysis and Design”, 2nd edition, John Wiley and sons

[8] Patent application WO2011 / 011307 A1, "System and Method for Improving MIMO Performance of Vehicular Based Wireless Communications"

[9] USPTO patent application US2009 / 0209212 A1, "Method and Apparatus for an Adaptive Multiple-input 15 Multiple-output (MIMO) Wireless Communications Systems"

[10] USPTO patent application US2010 / 007573 A1, "Multibeam Antenna"

[11] USPTO US2008 / 0225972 A1, "Method and System for Partitioning an Antenna Array and Applying Multiple-input Multiple-output and Beamforming Mechanisms" patent application

ES 2 426 321 A2

Claims (12)

1. Method for implementing a transmission mode of multiple inputs-multiple outputs, said MIMO transmission mode comprising using a certain number of antennas in a user equipment and a given number of antennas in a radio access node for communication between said user equipment and said radio access node, said communication carried out by means of a communication channel, said communication channel represented by a channel matrix, characterized in that it comprises:

-

using a range of said channel matrix associated with said communication channel less than the minimum of said certain number of antennas and said given number of antennas or if at least part of said coefficients of said channel matrix are correlated;

-

 de-correlate at least part of the coefficients of said channel matrix when said communication channel is degenerated using at least one additional antenna port on each of said radio access node and said user equipment in order to change radiation patterns associated with transmission layers of said communication allowing the use of an enhanced MIMO transmission mode; Y

-

 calculating, said radio access node, said range of said channel matrix according to channel quality indicators and range indicators sent from said user equipment to said radio access node.

2.

Method according to claim 1, wherein said communication is carried out in a mobile network of advanced long-term evolution.

3.

Method according to any of the preceding claims, wherein said at least one additional antenna port forms two networks with its two adjacent antennas of said given number of antennas or said certain number of antennas, each of said two networks associated with one of said transmission layers.

Four.

Method according to claim 3, wherein the distance between said at least one additional antenna port and each of said two adjacent antennas is at least half wavelength.

5.

Method according to claim 3 or 4, comprising having half the number of additional antenna ports with respect to said given number of antennas or said certain number of antennas, said additional antenna ports being intertwined between said given number of antennas or said determined number of antennas each of said additional antenna ports forming two networks with their two adjacent antennas, leaving an antenna of said given number of antennas or said determined number of antennas without pairing if there is an odd amount of said given number of antennas or said certain number antennas.

6.

Method according to claim 2, wherein said at least one additional antenna port forms a triangular network with two adjacent antennas of said certain number of antennas.

7.

Method according to claim 2, wherein said at least one additional antenna port is a panel antenna with pairs of cross-polarized elements when said antennas of said given number of antennas are linear cross-polarized antennas, which has an antenna of panel for each pair of said linear cross polarization antennas.

8.

Method according to any one of the preceding claims 3 to 6, which comprises feeding said at least one additional antenna port or each of said additional antenna ports with a linear combination of the signals that feed said two adjacent antennas, according to the following expression :

T3 = exp (i · φ1) · T1 + exp (i · φ2) · T2 where T3 is the signal that feeds said at least one additional antenna port or each of said additional antenna ports; T1 and T2 are said signals that feed said two adjacent antennas; φ 1 and φ 2 are the relative phases of said adjacent antennas; Y exp is the exponential function. ES 2 426 321 A2 9. Method according to claim 7, comprising feeding each of said cross-polarization elements of said panel antenna according to the following expressions: T3x = exp (i · φ 1) · T1 T3y = exp (i · φ 2) · T2 5 where T3x and T3y are the signals that feed each of said cross-polarization elements of said panel antenna; T1 and T2 are the signals that feed each of said linear cross polarization antennas into a pair of linear cross polarization antennas; 10 φ 1 and φ 2 are the relative phases of said linear cross polarization antennas in said pair of linear cross polarization antennas; Y exp is the exponential function. 10. Method according to any of the preceding claims 5 to 9, comprising choosing values for said relative phases in order to force nulls and / or peaks along certain directions in space in said radiation patterns. A method according to any of the preceding claims 5 to 10, comprising, in order to eliminate the degeneration of said communication channel, varying said relative phases when signals are transmitted from said radio access node or said user equipment, and / or adjusting said relative phases in signals received by each of said given number of antennas or said given number of antennas. A method according to any one of the preceding claims 8 to 11 which comprises performing, said radio access node, an adaptive search of the optimal values for said relative phases. 13. Method according to any of the preceding claims, which comprises informing said user equipment that said enhanced MIMO transmission mode is used to carry out said communication.